Midjourney V6 is in the Wild | A Guided Tour | AI Art Tutorial

hello the first release of mid Journey version 6 is in the wild and all mid Journey users have access to it on Discord today I'll focus on three topics for version six one showing off the much anticipated text capability of version six two reviewing the enhanced and more realistic results in images and three testing the very subtle and very strong functions of version six first let's do some brief housekeeping for the version 6 Alpha release since version 6 will not be a default model until early 2024 you can access version 6 Alpha through settings and select mid Journey model V6 as shown if you want temporary access to version 6 you may simply add-- v6.0 to the end of any prompt let's start with a quick overview of differences in version six according to Mid Journey prompting is significantly different it is said that we need to learn how to prompt for images in version six a key reason for this statement is that version 6 follows image prompts much more accurately version 6 also processes more tokens of very long prompts many more tokens than mid Journey could process in version 5.2 this means that prompt instructions previously ignored near the end of a long version 5.2 prompt can now be recognized in version six with new prompt accuracy mid Journey recommends stopping using junk tokens in prompts such as award-winning photorealistic obtain render 4K or 8K I'll include a link in the description area to a free prompt cleaner tool that I have developed for you it is easy to use and can remove many of these so-called junk tokens you'll also have a link to a file with all the prompts used in this video for our first topic let's enjoy the new text capab ability ever since Del 3 came out with rudimentary text functionality many mid Journey users have wanted text functionality I'll spend a little time on this new feature since it was much desired here's what mid Journey says about including text in your image the current version has a minor text drawing ability you must write your text within quotations or double quotes and you may wish to use-- style raw and or use lower stylized values such as 100 to improve text results here is a sample text prompt to render a SpongeBob character and a sign and here is the result very nice indeed next I have a prompt for a blue Hedgehog character with a white caption overlay I really like this result with the let's ride caption now a text prompt for a handwritten sticky note about washing the car I don't think I will forget this reminder I wondered how an attempt to write a message on a birthday cake might turn out it seems to me that this cake is good enough to eat I do like the colors of the icing as well I have noticed that shorter text seems to work better than longer text next we'll add the word bonjour to the sign of a Paris Street Cafe I found that adding the phrase single sign or single caption to prompts can help avoid duplicate text appearing in the image lastly I wanted an image of an old-fashioned Diner restaurant with a neon sign and this is the result I thought this was fantastic the sign was rendered wonderfully when using version six you should notice more detail in Renditions of humans and other subjects we start with four variations of a closeup of a desert dwelling girl with her camel in these images we see improved detail in the clothing material and in the patterns of clothing the skin looks more realistic and is not so airbrushed you can see finer strands of hair and eyelashes as well next we have four candid shots of couples on a boat ride in the ocean the requested candid aspect of the shot comes through in the images again the details of the hair and skin are striking and believable facial expressions are very natural as well you see realistic Renditions of wrinkles around the eyes and foreheads of individuals here are four variations of a close-up of a young woman working on an art project notice that in this early release of version six her hands are holding art tools correctly this was neither the case in version 5.2 nor in other AI art programs I also love how the AI placed art colors realistically smeared on the artists arms and hands we don't have in painting in version 6 Alpha yet there are nice things we can accomplish with both the very subtle and the very strong functions we'll begin with this photo of an Angelina Jolie lookalike wearing a lavender color leather jacket I think this is a fine result now I'll use the very subtle function to give her a gray colored leather jacket and sunglasses the very subtle function worked fine yet the colors on the graffiti wall are a bit washed out when compared to the original let's try this again with a very strong function with the same modified prompt in this version the colors are enhanced her pose is made more confident and the sunglass lenses darken all around the very strong image is better in appearance next I want a photo of a Captain America lookalike standing in front of a skyscraper I chose the pose and background of one of the results yet this pose lacks his shield as part of the photo I tried the very subtle function to get his shield into the pose that seemed to work very well for variation number three next I decided to put the skyscraper on fire and I knew the very strong function was more appropriate for this change I applied both the request for his shield and the skyscraper on fire and that yielded these four variations I like all of these variations and I was impressed with the kindor depth and accuracy of the results finally I wanted to see if I could do a type of in painting using very subtle and very strong Tim of the theoretically media channel inspired me to try this dysfunction I created a watercolor illustration of a mousse by a river which I thought came out nicely next I prompted for a subtle change from a moose by a river to a deer by a river mid Journey rendered a good-looking male deer or buck in place of the Moose now I wanted to see how a very strong function would change the result mid Journey was true to the prompt request and enhanced the colors as well I thought the result was very nice and I hope this inspires you to try the very subtle and very strong functions finally here is a list of features that are currently working in Alpha and those to come later in version six at the time of this video Drop mid Journey developers are already describing an imminent update including improvements in drawing text image coherence prompt accuracy and faster and cheaper UPS scaling in my opinion version 6 Alpha is already a nice step forward in upcoming videos I'll examine the responsiveness of version 6 to Artistic Styles camera angles and other photographic characteristics let me know your thoughts regarding mid Journey 6 and features you look forward to using thanks again for your support and feedback bye-bye

Channel: Carly AI

Views: 3,312

Keywords: ai art, midjourney, midjourney ai, midjourney v6, midjourney v6 prompts, midjourney v6 guide, midjourney v6 text, midjourney v6 rumors, midjourney 6, midjourney version 6, how to use midjourney ai, midjourney tutorial for beginners, how to use midjourney discord, how to use midjourney ai to create images, midjourney ai tutorial discord, ai art generator

Id: iBGOsEkKQvM

Length: 8min 25sec (505 seconds)

Published: Fri Dec 29 2023